Probably as good a time as ever to mention: In 2021 I'll be heading up the new Rust team at Facebook, focused on improving the Rust compiler and ecosystem for the Rust community, just like my colleague @nikomatsakis is doing at Amazon. :)
My big LLVM memcpy eliminator patch is up — 17% reduction in Rust memcpys now, with up to 42% possible in the future with more MIR optimization work. reviews.llvm.org/D140089
I'm getting more convinced that Rust code is generally going to end up faster than C++ code every day I work on optimizations.
Strong immutability and no-alias guarantees are a game-changer and we've only really begun to scratch the surface of what can be done.
A work-in-progress implementation of vector textures via sparse virtual texturing. This allows an SVG to be used as a texture for any mesh at effectively infinite resolution without using much GPU memory.
Amazing to me how many people deny that C++ has a problem with approachability of build systems and dependency management. It's total Stockholm syndrome.
colleges prioritized making people feel perfectly safe over everything else and produced a generation afraid to fail, and thus afraid to take risk, and thus on pace to accomplish extremely little
Landed a bunch of changes to optimize memcpys in LLVM 16. Rust standard library print() assembly code before and after (after I land a tiny fix to remove a workaround for bugs in old versions of LLVM in rustc):
